https://issues.dlang.org/show_bug.cgi?id=24021

RazvanN <razvan.nitu1305@gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |razvan.nitu1305@gmail.com

--- Comment #1 from RazvanN <razvan.nitu1305@gmail.com> ---
Note: Walter is against issuing warnings. The compiler should either accept or reject your code. A warning is something vague that programmers usually ignore (against what best practices advice). In the last 7 years there were exactly 0 warning added to the compiler.

With that being said, there is no way such code is ever going to be an error. That would mean that any function that receives an out parameter (or a pointer/ref parameter) would have to issue an error if used in an assert. That is very restrictive.

Even if it were issued as a measly warning, I don't really see the usefulness of it. Asserts have the purpose to test for a specific condition, in this particular case, that a function returns a specific value. Why would a warning be issued if a function receives a pointer parameter? What's the actionable item for this message? Stop using the assert?

--

https://issues.dlang.org/show_bug.cgi?id=24021

--- Comment #2 from RazvanN <razvan.nitu1305@gmail.com> ---
This could be implemented by a third party tool using dmd as a library. But I don't think it should be implemented in the compiler.

--

https://issues.dlang.org/show_bug.cgi?id=24021

Nick Treleaven <nick@geany.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |nick@geany.org

--- Comment #3 from Nick Treleaven <nick@geany.org> ---
Also asserts are used in unit tests where having side effects is often deliberate.

--

https://issues.dlang.org/show_bug.cgi?id=24021

mhh <maxhaton@gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |maxhaton@gmail.com

--- Comment #4 from mhh <maxhaton@gmail.com> ---
Pretty sure warnings have been added.

--

https://issues.dlang.org/show_bug.cgi?id=24021

Tim <tim.dlang@t-online.de> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |tim.dlang@t-online.de

--- Comment #5 from Tim <tim.dlang@t-online.de> ---
This seems to be a duplicate of issue 12028.

--

https://issues.dlang.org/show_bug.cgi?id=24021

Basile-z <b2.temp@gmx.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
                 CC|                            |b2.temp@gmx.com
         Resolution|---                         |DUPLICATE

--- Comment #6 from Basile-z <b2.temp@gmx.com> ---


*** This issue has been marked as a duplicate of issue 12028 ***

--

https://issues.dlang.org/show_bug.cgi?id=24021

--- Comment #7 from Grim Maple <grimmaple95@gmail.com> ---
With all due respect, this isn't a _duplicate_ of 12028. I'm not proposing to
disallow side effects, I'm proposing to notify user that their code was not
compiled.
I think this warning should be issued in release builds only, as they are the
only one affected. This will not affect unittests or bring other
inconveniences. Besides, maybe, rewriting code from `assert(foo())` to `auto
tmp = foo(); assert(tmp);`

--

https://issues.dlang.org/show_bug.cgi?id=24021

Basile-z <b2.temp@gmx.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
         Resolution|DUPLICATE                   |---

--- Comment #8 from Basile-z <b2.temp@gmx.com> ---
indeed, subtle difference.

--